Modelling introduction

The trac runtime is the Python library that trac models are written against. A model built with it runs unchanged in your IDE, in a notebook, or on the trac platform, where every run is recorded as a calculation contract — see Portable execution.

The runtime is open source, part of TRAC D.A.P., and is installed from PyPI:

pip install tracdap-runtime

Runtime concepts

The building blocks of a trac model: the model base class, how models are defined and launched, and the context available at runtime.

Modelling tutorials

Worked examples, from local setup and a first model through to chaining models, structured data, and importing and exporting data.

Model API

Reference for the runtime’s Python modules: the model API, exceptions, launch helpers and the metadata model.